On the Sadler Effect and Biases in Holocene Paleofire Records